Skip to content

fix(attention): correct XLA flash attention availability check and in… #13890

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
gagandhakrey wants to merge 2 commits into
base: main
Choose a base branch
from
+4 −4
Open

fix(attention): correct XLA flash attention availability check and in… #13890

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
4fb1551
fix(attention): correct XLA flash attention availability check and in…
gagandhakrey Jun 9, 2026
9b6c86a
Merge branch 'main' into fix/xla-flash-attention-guard
gagandhakrey Jun 9, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions src/diffusers/models/attention_processor.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -324,12 +324,12 @@ def set_use_xla_flash_attention(
Specify the partition specification if using SPMD. Otherwise None.
"""
if use_xla_flash_attention:
if not is_torch_xla_available:
raise "torch_xla is not available"
if not is_torch_xla_available():
raise ImportError("torch_xla is not available")
elif is_torch_xla_version("<", "2.3"):
raise "flash attention pallas kernel is supported from torch_xla version 2.3"
raise ImportError("flash attention pallas kernel is supported from torch_xla version 2.3")
elif is_spmd() and is_torch_xla_version("<", "2.4"):
raise "flash attention pallas kernel using SPMD is supported from torch_xla version 2.4"
raise ImportError("flash attention pallas kernel using SPMD is supported from torch_xla version 2.4")
else:
if is_flux:
processor = XLAFluxFlashAttnProcessor2_0(partition_spec)
Expand Down
Loading